Subduxion Responsible AI Policy

Effective October 6, 2025

At Subduxion, we believe AI should amplify human capability, not replace judgment. Subduxion is building a world model for precision manufacturing; Blake, our first application, uses AI to support manufacturing cost estimation. This policy sets out the principles we design and operate our AI systems against.

Our guiding principles


  • Transparency: We aim to make it clear when and how AI is used within our systems, and what a given output is based on.

  • Human oversight: Our AI systems are designed for human-in-the-loop review. The estimator or other user confirms, corrects or rejects a system proposal before it is acted on; that decision is the result.

  • Privacy and data protection: AI systems are operated in compliance with GDPR and applicable data laws. We do not use customer data to train models unless explicitly agreed in writing.

  • Fairness and non-discrimination: We aim to build AI systems that treat similar inputs consistently, and to review data and outputs for bias as part of our development process.

  • Explainability: We design our AI systems to be traceable: an output can be reviewed against the inputs and assumptions it rests on, so a person can check and correct it rather than simply accept it.

  • Security by design: Our AI systems, data pipelines and APIs follow the same principles defined in Subduxion’s Information Security Policy, including encryption in transit and at rest and access controls. Security is considered from design through deployment.

  • Compliance and accountability: We monitor emerging regulations, including the EU AI Act, and adapt our practices accordingly. Customers retain ownership of their data, and Subduxion remains responsible for the systems we operate.

  • Continuous improvement: We review our models, their performance and our governance approach as the product develops, so our practices keep pace with what the systems actually do.

Our promise

We do not build or support AI systems designed to deceive, manipulate, or cause harm. Our aim is technology that strengthens human decision-making and keeps the person using it in control of the outcome.

For questions or requests regarding Subduxion’s responsible AI practices, contact legal@subduxion.com.